The classical one-time pad

The one-time pad is not used in practice. One of the reasons for this is because the key needs to be as long as the message that is sent. Why is that? 

This is because of how the message is encoded: each character has a sequence associated to it, making the length of the sequence depend on the length of the message.

As Tim already suggested in the video, can you show on a piece of paper why sending two messages with the same one-time pad is a bad idea?

This is a cool exercise you need to do by yourself, but you have seen the idea in the video. If you use the same key to encode two different messages, Eve can perform a bitwise XOR operation between both encrypted messages. This way she will get a XOR operation between the two original messages, meaning she will be able to read both messages without ever knowing the key.

Can you think of a way we could use a shorter key multiple times?

The problem with the encoding we have just seen is the linear relation of encoding and decoding using a bitwise XOR operation. We will see soon that different protocols, such as RSA, do not encode using linear functions, which works very well!